A script using bootstrap techniques to calculate confidence intervals for parameter estimates from a 'dark' object.
Usage
BootDark(obj, R, graph, progress = F)
Value
Returns a list 'out'
out$time
times of observations
out$thrs
thresholds
out$opt
optimised parameter estimates
out$Mod
the name of the optimal model
out$Pn
number of parameters needed to describe the data
out$AIC
the AICc scores for the three models
out$fit
fitted values for the optimal parameter estimates
out$resd
residuals of the best fits
out$R2
the coefficient of determination
out$Bootstrap
bootstrap parameter estimates, 2.5%, 50% and 97.5%
out$weight
the reciprocal of the CI
out$valid
an indication whether the parameter estimate is valid
out$data
the source of the data
out$call
updates the call label on the object
Arguments
obj
A 'dark' object.
R
The number of repeats for the bootstrap calculations.
graph
A flag to indicate whether a figure should be drawn.
progress
A flag to indicate whether a progress bar should be drawn to the console. This might be preferred if using a large number of repeats.
Author
Jeremiah MF Kelly
Mumac Ltd, SK7 6NR, GB
Details
The script calculates bootstrap estimates of confidence intervals by sampling the residuals without replacement. The seven parameter model 'P7c' is always used. If 'P3' or 'P5c' have been found elsewhere to be a better fit then this will be confirmed by bootstrapping the 'P7c' model.
References
B. Efron. Bootstrap methods: another look at the jackknife. The Annals of Statistics, 7(1):1-26, 1979.
B. Efron. Nonparametric estimates of standard error: The jackknife, the bootstrap and other methods. Biometrika, 68(3):589, 1981.